Synthesia's Victor Riparbelli: Using AI To Replace Boring PowerPoints With Interactive Video

0:00The problem with fringe ideas is that most of them are wrong, but the few that are right are very, very valuable, right? For me, it was always thinking about, like, am I crazy or is there something here? And I think what always kept me afloat was the fact that there were select few people I really deeply, truly respected who was like, this is going to happen. Welcome to the Upstarts podcast, our weekly show where we talk to emerging startup founders about their upstart moment. Upstarts are challengers who punch above their weight to take on the status quo and improve the world, all while building a big business too.

0:34I'm your host, Alex Conrad, founder and editor of Upstarts Media. I'm excited to introduce our guest today, Victor Riporbelli, co-founder and CEO of Synthesia. This podcast is brought to you by Mercury, radically different banking that's already loved by 300 ,000 entrepreneurs. Thanks for joining, Victor. Thanks for having me. I was really excited to have you on because I feel like, you know, Synthesia has been around since 2017 when you co-founded it with three others. And I feel like more recently, you guys have started to really break out. You know, you announced$150 million in ARR, a$4 billion valuation.

1:07You work with over 90 % of the Fortune 100. You know, and I was looking, these are really big companies like IKEA, Merck, you know, Amazon. These are not small businesses. And so I thought, okay, here's a company that is going in the journey from Upstart to maybe this established, inspiring leader that other companies can learn from in real time. So that's why it was really exciting for me to talk to you kind of mid-moment, if that makes sense. For sure, yeah. And so you're from Denmark originally, right? Yeah. I read that you consider yourself maybe rebellious. You didn't love orders and schoolwork and stuff, but you did like World of Warcraft, right?

1:43yeah i think i think i've always just been more drawn to counterculture fringe ideas weird things no matter of the kind of music i listen to the kind of books that interest me what kind of music you can't you can't say that without saying what kind of music well i mean throughout my music is my my biggest interest i think you know throughout my life i've listened to like all the like weirdest subchanges you can imagine from like death metal to minimal techno which is probably like what i listen to the most but i love crate digging finding vinyls for one in some random record store that no one else has heard.

2:11I think I'm just very, I kind of like that more than I like the more kind of conventional ideas and media. I find that life's more interesting at the fringes. And I think when it comes to business, I mean, there's many ways being good at business, but I think one of the things that most entrepreneurs have to be pretty good at is like looking around corners and seeing trends, right? And most trends start as fringe ideas. Ours definitely did. The idea of like making video with AI without cameras, actors and studios, like in 2017 was a very fringe idea. But even like lots of other things, like Airbnb, right?

2:43Also like totally weird fringe idea that you would like couch surf on like random people's houses. Like being a good entrepreneur, I think is like spotting those things that are indicative of what's going to happen in the future. And the hard thing is that there's a lot of fringe ideas and most of them go nowhere, but a few of them actually end up being ideas that transform the world in some way, shape or form. Did you think growing up that you wanted to be an entrepreneur or startup something that you studied or admired? or sort of how did you go from music lover to, you know, making websites, doing, you know, some of the early startup work you did before your company?

3:17Not at all. Like it really just kind of happened. I think when I was younger, I didn't really know what I wanted to do. If you were to ask me back then, I would have probably done something with music, something creative. I went to business school like early on and hated it. I thought it was like really boring. I didn't like the people. Was that in Denmark? That was in Denmark. Okay. I thought it was boring. I thought I didn't like the people. It was just like, not for me. The turning point kind of became that I started doing these websites, trying to make some money. My dad was just like, you're really good at computers and you're at your computer all day.

3:50You've got to be able to make some money with that. Obviously, even back then, right? You could have sold World of Warcraft characters. I did. I did. I actually had bots running around harvesting gold and leveling up the characters that I could then sell. But that was when I was like 12, 13. This is probably like five years later. That was your first hustle. Yeah, I love that. That was my first hustle. And then later on it became, you know, my dad was just like, why don't you, you know, instead of just like hacking around your own, do something with other people. I was like, yeah, that's probably a good idea.

4:17So I started doing website, e-commerce stores, and I think just arbitrage and Shopify, which is interesting because like now I think there's like a new arbitrage and vibe coding and all the cool things you can do without having to be a developer, right? Back then you could set up an e-commerce store without being a developer and people would look at me like, really, you're going to make a web shop for me for$500 when everyone else is charging$10 ,000? Who are the customers for this? It's like the local store in the city I grew up in who sold tennis equipment, for example. Really, really small things like that.

4:47And after doing that for a while, my dad was like, why don't you just study something with computers? And I was like, that's actually a pretty good idea. Why don't I do that? And I think what it gave me, this thing of creating websites, was there was a mix of, of course, it's business, right? But it's also about design. There's a human element to it. once I started studying and I kind of I kind of got thinking of like making e-commerce stores is like for small businesses was fun and I made a lot of money on it but it wasn't really that intellectually stimulating so I started getting like well who's the the guys who made Shopify who actually did that right yeah and I got into kind of the startup ecosystem and what I think I found there was something that felt a lot less like businessy as in like how I thought of like businessy people and like going to business school it was a lot more of like you know creative people there was that rebellious energy right of like i think you have to be when you're if you're an entrepreneur you have to think that you can you know come out of nowhere and and capture a big slice of the market and i like that it's usually about like replacing an incumbent like there's that kind of rebellious energy to it and i just like saying upstart energy even exactly and that really attracted me because i think combined like the creative things in me with this also you know i think my worst nightmare back then was like if i ended up becoming like a mckinsey consultant or like working in a bank or doing something like that which was just to me be the most like soul destroying thing i'd possibly spend my life on but this kind of had i think i felt i was good at it and i think that's really important when you choose what to do in life i felt i was much better than i was at the music thing to be completely honest with you and i think it just hit a lot of things where I kind of felt like in flow when I was doing this when I was working in startups I loved it I would come home and I would naturally just keep working not because I had to but because I found it generally like very stimulating I think that's how I you know got into this world and then I think my studies kind of like took a bit of a backseat didn't really interest me that much I generally have a very hard time learning things that I'm not like genuinely interested in and even though I think my so I have like a split I started the IT University of Copenhagen.

6:51And I did a split degree. You would learn both some elements of business, some elements of coding and developing, elements of design. And I just felt like most of the things they were teaching us was outdated, not very interesting. So I did my bachelor's and it was okay, but I was working full-time the entire time and just fell in love with startups and since then figured out wanting to build my own thing. So how do you get from student in Denmark to London, getting to know the startup ecosystem here, maybe thinking about what your big idea is going to be? Yeah. So in Denmark, I was working in the Danish ecosystem.

7:24And as you do today and back then, right, like you look a lot to Silicon Valley, you read all the books, you follow the people on Twitter. And I always felt like I was sort of immersed in that world. But physically, I was still in Denmark, right? As part of my degree, I had one semester where you could go out and go somewhere else. And I got the opportunity to go to Stanford. And that was a very transformational life experience for me, because as much as I love the Danish startup ecosystem, it's very different than what I met in the US. The US, people think way bigger. People want to build startups around crazy ideas, like frontier tech.

7:57Whereas in Denmark, I felt like it was a lot of like, let me build a little accounting tool for small Danish businesses. Let me build this business process optimist. On a personal level, it just didn't speak to me very much. I loved the ambition that I felt in the US. And so I kind of actually wanted to go to the US, but pretty difficult without a visa. I didn't want to go work for a big company because I knew I wanted to start my own thing. And this is like pre-Brexit. So anyone could just, you know, move to the UK, had a bunch of friends here. So it's actually a very certain place that I just ended up here.

8:25But I felt like, you know, I definitely can't do it in Denmark. This is probably like a good kind of second and best place to go if I couldn't get to the US. And so I went here and I was working with this professor of machine learning that I met for a year doing a bunch of like consulting freelance gigs. We worked for the government, a subsidiary of the government over here. who were trying to figure out how to make London the best place in the world to build virtual reality and AR startups. This is in 16. So that was like the hype cycle of the time. Right. That was a great gig because I got to write a bunch of reports for them.

8:56I got to set up what's called a volumetric studio, which is this Hollywood visual effects, like 82 cameras and kind of an array. You'd go in the middle, you'd like do a dance or something and we'd get a 3D asset out of it, which made it easier to create virtual reality content. So I got very immersed in that world, met a lot of people and I kind of realized, I think, that I love VR on a personal level, but I just felt like it's too clunky, it's too difficult to make content. Like the time is not now for VR. It's not gonna, it's gonna be, it's gonna pass. But I got introduced to a lot of people who are working on content creation using kind of early machine learning techniques.

9:29And one of those people was Professor Matthias Niesner, who ended up becoming my co-founder. He did a research paper called Face to Face, which was like one of the first, some would call it like the first deepfake paper, right? but the first paper showing like a neural network actually generating video that looked very realistic. And outside of work, I've always been, you know, I've been making music. I've done like 3D rendering, played around with Photoshop, never as like an expert as a job, but just as like a personal interest. I think I saw that. I was just like, I knew how long it would take to create the clips he was showing me in a Hollywood visual effects studio, right?

10:03Like what he's showing me working in actually that worked in real time back then, like this is going to take three months and 10 people in Hollywood studio to do this. There's something very interesting about this. And that's kind of how the idea around Synthesia started. There was these two big ideas that I kind of wrote down as our kind of founding thesis. The first one being that, you know, AI is not just going to be able to help us analyze and predict patterns in data, which is like what everyone was doing back then, more or less. It's going to be able to eventually generate data. And when you can generate data, that means you can begin to generate content, video, audio, music, whatever.

10:37and when that happens right the the way we're going to create content is going to change very significantly and it's going to be as drastic as from when we used typewriters to when we went on keyboards right in like the creation of the content distribution of the content is going to change everything and it's going to drive the marginal cost of creating content to zero not just in dollars but in time and skill required and that is that felt like a big business opportunity and i think we look at where the world is right now we're bang in the middle of it right these technologies are now good enough to create a lot of different types of content.

11:06We're focusing on business, enterprise, video, but we're starting to see that for entertainment, this is also becoming really, really good. Music videos, as we're recording this, there's a new model from ByteDance called SeatDance 2.0, which looks to maybe be the model that truly breaks through and is good enough for cinematic content. All this stuff's happening, and that's got to be very transformational, right? and that was kind of the thesis back then the second part of it was that what always happens when we as humans invent new media technologies right is that we invent new media formats which are native to those technologies and that was i've always found that to be the philosophically like most interesting part about what we're doing we're about to launch our second product now which is video you can talk to and essentially kind of like if you had to reimagine video in the age of ai what would it look like right knowing that we have llms we have offline video models of very high quality we have real-time video models of a bit lower quality we have all these different technologies right so what can a video look like and it's going to look more like a mix between a video and a computer game and we could talk interactive interactive real-time video let's definitely talk about that in a bit but i want to i don't want to get too far ahead obviously you're super passionate about this all these years in was there ever a moment of doubt as you're starting that hey we're we're crazy or this might not work you know i i know you were younger in 2017 you're working with two professors and another co-founder.

12:25What was your feeling at the time in terms of the degree of difficulty, what it would be like to actually build a real business as you're getting going? I'm very much an optimist in life, and I think that has helped me a lot throughout this journey because you could argue that we started too early. You could also argue we were right on time, given that things have panned out fairly well. But the first few years, the technology basically didn't really work. It was impossible to make... So the first three years, we raised the first round of funding from Mark Cuban. Incredibly difficult round to raise because back then, everyone wanted to invest in fintech companies and the kind of current hype thing.

13:05We were telling people in 10 years, you can make Hollywood films on laptops with just imagination. And that's going to be a big business. People are going like, cute idea, but not going to fund you. So I saw that I think you were rejected by almost 100 investors, but you're not the first founder on this show to talk about 100 investors. So it seems like maybe it's a sign of a good company that people don't get it right away. What was the biggest hurdle for people to understand it? And what did you do to convince Mark Cuban to invest and kind of, you know, get things rolling? I think a bit with this thing about like rebelliousness or like fringe ideas, right?

13:37The problem with fringe ideas is that most of them are wrong. But the few that are right are very, very valuable, right? And so I think for me, it was always thinking about like, am I crazy? Or is there something here? And I think what always kept me afloat was the fact that there were select few people I really deeply, truly respected, who was like, this is going to happen. Like, you're right. The timeline, you know, we can discuss, but it's going to happen. And it's not crazy. Even though 99 % of people would tell me the opposite, like having a few of those people like support you that I think was very helpful.

14:12The learning for me was that the way we view the world, the kind of Overton window in some sense, if you will, in how VCs invest changes all the time, right? Like obviously now everyone is investing in AI, everybody wants agents, nobody wants SaaS any longer. That's like a total no-go. And when you're trying to fund something which is not in that window of like the current narrative, it's really difficult. And it's very, very hard to convince most VCs of something, of like a vision that's not within that, right? So back then we would go out and the starting point would, of course, be like, do you think this is going to happen in 10 years?

14:46If they didn't agree, if they think that's true, then they're going to start evaluating you and the team and your strategy of how you're going to do it. But if you can't convince them of the first part of it, it's very hard. And I think what I've learned since then, both my own angel investing and our journey back then, is if you're doing one of these things, you need to find someone who shares your worldview. And Mark Cuban did. He had actually implemented this paper face-to-face himself at home. He was totally convinced that AI would change everything about media production. So we didn't have to really discuss that much with him.

15:14We kind of relied on that from the beginning. And he was much more interested, what's your strategy? How are you going to sequence it? Who are you? We're the founders. And I think the learning there is just, it is, of course, possible to convince someone of something they don't already believe, but it's very hard. And usually your time is actually better spent on trying to find people that already share that worldview. When I launched Upstarts Media last year, I had big ambitions and a lot to figure out. I knew I needed a bank that could move at my speed and grow with me too. With Mercury, I was able to get started in minutes without leaving my desk.

15:49Mercury made it easy to send and receive wires and manage the cash flow of our growing business from day one. Now, I can spend less time with busy work and more time with the founders I like to cover. Building a media startup comes with new challenges every day. With Mercury, banking is one I don't need to worry about. You don't accept outdated tools like floppy disks and fax machines anywhere else in your work and life. Why accept them for something as important as your money? Visit mercury.com to learn more and apply online in minutes, like I did. Mercury is a fintech company, not an FDIC-insured bank.

16:20Banking services provided through Choice Financial Group and Column N.A., members FDIC. Was that cold start or zero to one moment the moment that maybe felt the most challenging? Or have there been other moments that were sort of key inflection points now as you look back? On the show, we call them an upstart moment. And I'm curious when you think what your real forged-in-fire moment was, what comes to mind? I think there's a lot, right? Of course, getting the first round of funding in is pivotal because that's what enabled us to get started. We then spent the first few years on building an AI dubbing product, a technology that now is pretty widespread because technology is mature enough.

16:56It just works very easily. Back then, it didn't really work. It was very handheld. Creating a 30-second clip would take two PhDs, and it was just not really a scalable business. And so the kind of pivotal moment there was figuring out, actually, we're building for the wrong people. We were building for video professionals, Hollywood studios, advertising agencies, the most obvious people to buy AI video technology. And I think generally with new technologies, it's very often that it's not the most obvious group of people who adapt it first. They actually adopt it last. And I think it's a bit the same with Hollywood and AI video.

17:25I think Hollywood's going to be the last ones to really adopt AI video. but we're building for them and we figured out that actually the much more interesting market was people who work corporate jobs who had some important message to share with their customers or with their employees whatever and they could only do that in text or powerpoint because they wouldn't have budget to create videos they couldn't get it from their boss the content they were creating was not like important or high value enough to warrant a video if they did make a video it'd be out of date by the time it was edited and didn't know how to make the video themselves but they were telling us like look everybody wants to watch videos even back in 2020 right that trend was tiktok was on the rise obviously youtube like on a on a roll and that has only continued since then they're just like can you help me make video in some way shape or form and the moment we started focusing on them that was very pivotal that's how we invented the avatar technology because we figured out that actually what these people need they don't need cinematic scenes of crazy stuff happening with the Hollywood films.

18:23They just need a person talking to the camera, which was a much more tractable problem back then where the technology was still not very mature. That was definitely one of those pivotal moments. And I think essentially a way to look at it is we went from focusing on people who create a low volume of videos of very high quality, I think your advertisements, films, etc., to essentially very high volume videos, but at much lower quality. If you look at most corporate video, it's a different quality level that's expected than in the kind of storytelling creative world. And yeah, I think that was definitely one of those moments.

18:59Okay, but you've already established that you're a bit of a skeptic about authority, a free thinker, musical guy. Did it take a little bit of a rethinking or a reassessment of your own approach to get excited about enterprise? Or were you just so excited that people needed the product that it felt as exciting to build for them as it might the Hollywood types? Yeah. I mean, I think, you know, I did set out in the beginning to do more like the Hollywood film kind of stuff, but I actually equally excited what we do today. I think for me, the product building is a creative exercise. And what we're doing is not necessarily helping you create like, I mean, I would say the quality of the content that people have created over the years is massively, it's getting better and better.

19:42I had to ask you your most recent launch video. I was like, which parts are the AI victor and which parts are the human victor, right? It's getting really good, right? I think what we're doing now is like, I mean, we're less focused on giving tools like directors and professionals. But we're helping a lot of people who are creators, right, just in a different context, make videos. And I think that's really fun. And I think it's also fun because rather than building for video experts, you're trying to bring new people, like in our case, PowerPoint users to video. And that creates a lot of interesting design questions of like, we've always anchored like our thinking in PowerPoint and how people use PowerPoint.

20:18But of course, you won't get stuck in that paradigm. We still want to push you forward and invent new ways of using these tools. Right. I think that's been really fun. And I think now we're also starting to see, you know, a lot of the things you can do is like there's no, very soon it's not going to be any kind of barriers for a training video on a big enterprise to look like a Steven Spielberg film. You know, like that's just going to be limited by your imagination. The technologies and the AI models are going to be so good that you can actually create very high quality, funny, interesting, beautiful cinematic content that works in enterprise context.

20:50And that is really fun, actually. And seeing how our customers, when they pick up our tools, how creative they become when you let them is a magical feeling. That's really interesting to me because my first thought is, do office workers want the Spielberg quality video? If it's a learning and development video or a training video, does a higher quality lead to better outcomes? Does it make them noticeably happier? Or how utilitarian is the end result? So it is, of course, very utilitarian. But I think also the reason a lot of this video is utilitarian is because it's really the only option you have, right?

21:27Like if you're creating a simple, like a training video, for example, you don't have three months or a big budget to make it really interesting and cool. but what is i think universally true is that all of us we go through our days we consume a lot of information right like you consume text you consume video podcast whatever and not most of that you're probably not going to remember right like so standing out in that sea of information is really important and what we've seen over the last five years is just having a video versus a text document already makes you stand up much more increases information retention and engagement that content by a lot, which is very, very, very visible.

22:03But the better that content is, the more it will stick. So to give you just like one fun example, one of our customers is a big like chemical manufacturing plant and they have this factory floor where you cannot bring in chocolate. If you bring in chocolate, there's some chemical reaction or something. So that's like very, very important that no one brings chocolate into this facility. So this would be my wife's worst nightmare if she was not allowed to have chocolate. It's probably a lot of people's nightmare. So they've created a lot of videos with us, you know, throughout the years explaining why this is the case.

22:29they have posters everywhere and then we launched this new feature where you can prompt like anything you want basically not just an avatar talking but if you want you know someone jumping out of a plane like anything you can think of and then this team was like testing out the feature and one of them created this video of like a huge chocolate bar walking into the facility and exploding right and it looks like when you look at it like wow they spent 50k with some advertising agency to create a really cool animation but obviously this is just someone spending five minutes like prompting around and that got like really popular because it's really funny.

22:59And so it stands out much more than just like a sign on the one end of the scale saying it can't be chocolate in here, right? Maybe that's a novelty factor, but I think what will happen is just that the table stakes for how you communicate to people is going to be that it's going to have to stand out. It's going to have to be fun because your attention span is for content that you don't find interesting or it's not well made is low. And I think we're already seeing that with this, you know, stat about like our attention span kind of like decreasing very significantly, which a lot of people think is a negative thing.

23:31And I'm not saying it's not, but if I were to kind of challenge that narrative, right? I think there is, I think you could put forward an idea that maybe what's happening right now, especially the younger generation, is because we're all in these social feats all the time, we're kind of like the best and most engaging content will float to the top. We get used to content that is highly engaging, right? And some of that is, of course, AI slop and content we can deem crappy. But even for myself, like when I'm learning things, I go on YouTube to learn something new about music or AI or whatever, right?

24:00Some of that is short form content on TikTok, like a form of video explaining a content really rapidly. Some of it is a 30-minute YouTube video, but I'm used to now consuming really good content, right? So for me to pick up an academic book with 200 pages to try and learn something that I could learn through a 20-minute YouTube video that's really well made, made by someone who's young, they're funny, they're using lots of visuals to explain it. like is it a is it is it necessarily like a bad thing that my attention span isn't reading a 200 page book i think we definitely argue that it is but i do think there's an element of this which is just we're getting used to really high quality engaging content which wasn't the case 20 years ago right there was like five books you could read and you would just have to like sit down and read it and so i think there's something interesting there and i think no matter if we think that's a good or bad thing it's definitely true that that's the case right and so to stand out not just in being internal is maybe a workshop but like external facing communication you just gonna you have to stand out right but so you're walking into a big company like a merc or an ikea do they get that as well immediately or are they skeptical about trusting a startup to do this sort of thing how did you guys get in the door and and sort of start to score these really big customers so i think one of the things that is very good about building an ai right now is the just the level of interest.